> For clean Markdown of any page, append .md to the page URL.
> For a complete documentation index, see https://docs.beatapi.io/llms.txt.
> For AI client integration (Claude Code, Cursor, etc.), connect to the MCP server at https://docs.beatapi.io/_mcp/server.

# Has my AI read the article just because search found its link?

> POST /v1/web/search returns candidate URLs and short snippets only (news results have no snippet); pass selected URLs to POST /v1/web/read and verify results[].content before citing.

No. A hit from `POST /v1/web/search` means only that the URL appeared in search results, not that the page body was read. Search is for discovery, not evidence; pass chosen URLs to `POST /v1/web/read` and verify `results[].content` before quoting dates, versions, or policies.

## Follow a two-step Search-then-Read workflow

1. **Shortlist with Search**: Use `POST /v1/web/search` (`max_results` `1–10`, default `5`) to locate candidate URLs. Mark any fact seen solely in a search snippet as unverified, and add headline-only `type: "news"` results (which carry no `snippet` field) to the reading queue.
2. **Verify with Read**: Pass the selected URLs to `POST /v1/web/read` (`format`: `markdown` default or `text`; `max_chars`: `500–100000`, default `20000`; optional `query` up to 400 characters to return only relevant passages).

```bash
curl -sS https://api.beatapi.io/v1/web/read \
  -H "Authorization: Bearer $BEATAPI_API_KEY" \
  -H 'Content-Type: application/json' \
  -d '{"urls":["https://example.com/article"],"format":"markdown","max_chars":8000}'
```

For high-stakes topics such as API retirements, policy changes, finance, or health, always read the primary source page before finalizing recommendations.

## Contracts and related questions

Before running a request, inspect the contract for free: `data:web.search`, `data:web.read`, `data:web.map`.

[Back to Web Search questions](/faq/web).